Open Menu
AllLocalCommunitiesAbout
lotide
AllLocalCommunitiesAbout
Login

To Curb Online Sexual Abuse of Children, Experts Look to AI: Researchers in Norway and the U.S. are training artificial intelligence to address cybergrooming. Will it work?

⁨38⁩ ⁨likes⁩

Submitted ⁨⁨1⁩ ⁨month⁩ ago⁩ by ⁨Tea@programming.dev⁩ to ⁨technology@lemmy.world⁩

https://undark.org/2025/03/31/cybergrooming-children-ai/

source

Comments

Sort:hotnewtop
  • Jomega@lemmy.world ⁨1⁩ ⁨month⁩ ago

    Get ready for a lot of false positives.

    source
    • simple@lemm.ee ⁨1⁩ ⁨month⁩ ago

      If this is implemented right it should flag accounts so human reviewers can follow up on it, not take action on its own.

      source
      • Inucune@lemmy.world ⁨1⁩ ⁨month⁩ ago

        Even still, the ‘flag’ could be enough damning evidence for some people to take action. We’re in the cultural ‘guilty until proven innocent’ territory, where a mere accusation ruins lives.

        source
        • -> View More Comments
  • DemBoSain@midwest.social ⁨1⁩ ⁨month⁩ ago

    nicknamed SERI for “Stop CybERgroomIng.”

    Maybe we should have an AI that generates proper acronyms.

    source
    • MonkderVierte@lemmy.ml ⁨1⁩ ⁨month⁩ ago

      How about SERN?

      source
  • General_Effort@lemmy.world ⁨1⁩ ⁨month⁩ ago

    I guess most people don’t get how terrifyingly dystopian this is.

    In the EU, there is a serious push to make this mandatory.

    source
  • devfuuu@lemmy.world [bot] ⁨1⁩ ⁨month⁩ ago

    Get ready when the ai bota start behaving like chidren to bait and create a relationship with people.

    source
    • dustyData@lemmy.world ⁨1⁩ ⁨month⁩ ago

      This has already happened. There was a news article about a police force who used AI to bait groomers. This is further automation in something that’s already being done.

      source
  • sucrerey@lemmy.world ⁨1⁩ ⁨month⁩ ago

    weird question: if this worked, couldnt the same dataset be used to create a very skillful AI cybergroomer chatbot if it fell into the wrong hands?

    source